640 THE TEST. 



and of a transverse axis in Echinometraclae, owing to the obliquity of the 

 test, is an embryonic feature retained in the Echinometradae, and which 

 exists in all Echinoderms in their Platens stage. In one genus of the Echi- 

 nometradae (PI. X".f. 2) the short axis makes an acute angle with the 

 longer axis; in another it is the reverse (PL III'./, i). In Stomopneustes 

 the obliquity is so slight as to l»e imperceptible in many specimens. 



Loven has made an exceedingly ingenious analysis of the test of the three 

 suborders of Echini, in which he attempts, on entirely new grounds, to deter- 

 mine the position of the madreporic body as the right anterior one in the 

 regular Echini. From what I have attempted to show of the position of the 

 madreporic body with reference to an axis determined by the course of the 

 alimentary canal. I think Loven's views cannot lie sustained. The homology 

 he traces in the arrangement of the peristomal plates is one which is due to 

 the nature of the case. Any series of plates with an angular suture alternat- 

 ing on each side of a median line, and increasing in number at the apical 

 extremity in a spiral order, must necessarily consist of alternately larger and 

 smaller ones, and the unequal rate of growth, as well as its complete inde- 

 pendence in the ambulacra] and interambulacral plates, to which I called 

 attention in the Preliminary Report on the Florida Deep-Sea. Echini, seems 

 to me to form strong reasons for objecting to the general laws Loven at- 

 tempts to deduce from them. The Clypeastridae, standing, as they do, inter- 

 mediately between the regular Echini and the Spatangoids. do not come 

 strictly within the scope of the ordinal arrangement by which he has so 

 ingeniously connected the Spatangoids with the Desmosticha ; as in them 

 (the Clypeastridae) neither the connection of the madreporic body with the 

 right anterior genital plates ever exists, nor do we find the interambulacral 

 spaces in many genera reaching the actinostome, which is entirely sur- 

 rounded by ambulacral plates, while the interambulacral plates commence 

 with a single irregularly shaped plate, entirely outside the peristomal ring of 

 the ambulacral plates. 



